Greenplum6 數據庫數據庫學習_基本語法

時間 2021-04-14

標籤 sql 數據庫 app dom ide post this postgresql code 繼承欄目 SQL 简体版

原文原文鏈接

Greenplum6 數據庫

Greenplum6 數據庫分佈

1. hash分佈
2. 隨機分佈
3. 複製分佈

基本語法介紹

1. 獲取語法

[gpadmin@mdw logs]$ psql
psql (9.4.24)
Type "help" for help.

postgres=# \h
Available help:
  ABORT                            ALTER TEXT SEARCH TEMPLATE       CREATE RESOURCE GROUP            DROP FUNCTION                    LOAD
  ALTER AGGREGATE                  ALTER TRIGGER                    CREATE RESOURCE QUEUE            DROP GROUP                       LOCK
  ALTER COLLATION                  ALTER TYPE                       CREATE ROLE                      DROP INDEX                       MOVE
  ALTER CONVERSION                 ALTER USER                       CREATE RULE                      DROP LANGUAGE                    NOTIFY
  ALTER DATABASE                   ALTER USER MAPPING               CREATE SCHEMA                    DROP MATERIALIZED VIEW           PREPARE
  ALTER DEFAULT PRIVILEGES         ALTER VIEW                       CREATE SEQUENCE                  DROP OPERATOR                    PREPARE TRANSACTION
  ALTER DOMAIN                     ANALYZE                          CREATE SERVER                    DROP OPERATOR CLASS              REASSIGN OWNED
  ALTER EVENT TRIGGER              BEGIN                            CREATE TABLE                     DROP OPERATOR FAMILY             REFRESH MATERIALIZED VIEW
  ALTER EXTENSION                  CHECKPOINT                       CREATE TABLE AS                  DROP OWNED                       REINDEX
  ALTER EXTERNAL TABLE             CLOSE                            CREATE TABLESPACE                DROP PROTOCOL                    RELEASE SAVEPOINT
  ALTER FOREIGN DATA WRAPPER       CLUSTER                          CREATE TEXT SEARCH CONFIGURATION DROP RESOURCE GROUP              RESET
  ALTER FOREIGN TABLE              COMMENT                          CREATE TEXT SEARCH DICTIONARY    DROP RESOURCE QUEUE              REVOKE
  ALTER FUNCTION                   COMMIT                           CREATE TEXT SEARCH PARSER        DROP ROLE                        ROLLBACK
  ALTER GROUP                      COMMIT PREPARED                  CREATE TEXT SEARCH TEMPLATE      DROP RULE                        ROLLBACK PREPARED
  ALTER INDEX                      COPY                             CREATE TRIGGER                   DROP SCHEMA                      ROLLBACK TO SAVEPOINT
  ALTER LANGUAGE                   CREATE AGGREGATE                 CREATE TYPE                      DROP SEQUENCE                    SAVEPOINT
  ALTER LARGE OBJECT               CREATE CAST                      CREATE USER                      DROP SERVER                      SECURITY LABEL
  ALTER MATERIALIZED VIEW          CREATE COLLATION                 CREATE USER MAPPING              DROP TABLE                       SELECT
  ALTER OPERATOR                   CREATE CONVERSION                CREATE VIEW                      DROP TABLESPACE                  SELECT INTO
  ALTER OPERATOR CLASS             CREATE DATABASE                  DEALLOCATE                       DROP TEXT SEARCH CONFIGURATION   SET
  ALTER OPERATOR FAMILY            CREATE DOMAIN                    DECLARE                          DROP TEXT SEARCH DICTIONARY      SET CONSTRAINTS
  ALTER PROTOCOL                   CREATE EVENT TRIGGER             DELETE                           DROP TEXT SEARCH PARSER          SET ROLE
  ALTER RESOURCE GROUP             CREATE EXTENSION                 DISCARD                          DROP TEXT SEARCH TEMPLATE        SET SESSION AUTHORIZATION
  ALTER RESOURCE QUEUE             CREATE EXTERNAL TABLE            DO                               DROP TRIGGER                     SET TRANSACTION
  ALTER ROLE                       CREATE FOREIGN DATA WRAPPER      DROP AGGREGATE                   DROP TYPE                        SHOW
  ALTER RULE                       CREATE FOREIGN TABLE             DROP CAST                        DROP USER                        START TRANSACTION
  ALTER SCHEMA                     CREATE FUNCTION                  DROP COLLATION                   DROP USER MAPPING                TABLE
  ALTER SEQUENCE                   CREATE GROUP                     DROP CONVERSION                  DROP VIEW                        TRUNCATE
  ALTER SERVER                     CREATE INDEX                     DROP DATABASE                    END                              UNLISTEN
  ALTER SYSTEM                     CREATE LANGUAGE                  DROP DOMAIN                      EXECUTE                          UPDATE
  ALTER TABLE                      CREATE MATERIALIZED VIEW         DROP EVENT TRIGGER               EXPLAIN                          VACUUM
  ALTER TABLESPACE                 CREATE OPERATOR                  DROP EXTENSION                   FETCH                            VALUES
  ALTER TEXT SEARCH CONFIGURATION  CREATE OPERATOR CLASS            DROP EXTERNAL TABLE              GRANT                            WITH
  ALTER TEXT SEARCH DICTIONARY     CREATE OPERATOR FAMILY           DROP FOREIGN DATA WRAPPER        INSERT                           
  ALTER TEXT SEARCH PARSER         CREATE PROTOCOL                  DROP FOREIGN TABLE               LISTEN                           
postgres=#

2. 建立數據庫

[gpadmin@mdw logs]$ createdb testDB -E utf-8
[gpadmin@mdw logs]$ psql -h 10.10.10.101 -p 5432 -d testDB -U gpadmin
psql (9.4.24)
Type "help" for help.

testDB=# \q
[gpadmin@mdw logs]$ export PGDATABASE=testDB
[gpadmin@mdw logs]$ psql
psql (9.4.24)
Type "help" for help.

testDB=#

3. 建表語句

-  GreenPlum中建立表須要指定表的分佈鍵。
-  若是表須要用某個字段分區，能夠經過partition by 將表建成分區表。
-  能夠使用like操做建立與like的表同樣結構的表，功能相似create table t1 as select * from t2 limit 0。
-  能夠使用inherits實現表的繼承，具體實現參考postgresql文檔。

--語法查詢
\h create table
--建立表
create table test001(id int,name varchar(128)); --id 爲分佈鍵
create table test002(id int,name varchar(128)) distributed by (id); --同上

testDB=#  create table test001(id int,name varchar(128));
NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'id' as the Greenplum Database data distribution key for this table.
HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
CREATE TABLE
testDB=# create table test002(id int,name varchar(128)) distributed by (id); 
CREATE TABLE
testDB=# 

create table test003(id int,name varchar(128)) distributed by (id,name) --多個分佈鍵
testDB=# create table test003(id int,name varchar(128)) distributed by (id,name);
CREATE TABLE

create table test004(id int,name varchar(128)) distributed randomly; --隨機分佈鍵
testDB=# create table test004(id int,name varchar(128)) distributed randomly; 
CREATE TABLE

create table test005(id int primary key ,name varchar(128));
create table test006(id int unique ,name varchar(128));

testDB=# create table test005(id int primary key ,name varchar(128));
CREATE TABLE
testDB=# create table test006(id int unique ,name varchar(128));
CREATE TABLE
testDB=# 

採用隨機分佈策略的表默認將主鍵，或者惟一鍵做爲分佈鍵，由於每一個Segment都是一個單一的數據庫，單個數據庫能夠確保惟一性，多個數據庫節點就沒法保證全局的跨庫惟一性，故只能按照惟一鍵分佈，同一個值的數據都在一個節點上，以此來保證惟一性。

--若是指定的分佈鍵與主鍵盤不同，那麼分佈鍵會被更改成主鍵。在greenplum6 中這句話貌似顯示不兼容，以下：

create table test007(id int unique,name varchar(128)) distributed by (id,name);

testDB=# create table test007(id int unique,name varchar(128)) distributed by (id,name);
ERROR:  UNIQUE constraint and DISTRIBUTED BY definitions are incompatible
HINT:  When there is both a UNIQUE constraint and a DISTRIBUTED BY clause, the DISTRIBUTED BY clause must be a subset of the UNIQUE constraint.
testDB=# 
按照它的要求進行修改，咱們distribute by 修改爲id
testDB=# create table test007(id int unique,name varchar(128)) distributed by (id);
CREATE TABLE
testDB=# 

---建立如出一轍的 表
create table test_like (like test001);

--使用like 建立表的時候，只是表結構會跟原表一摸同樣，表的特殊屬性並不會同樣，例如壓縮，只增（appendonly）等屬性，若是不指定分佈鍵，默認分佈鍵與原表一致。

testDB=# create table test_like (like test001);
NOTICE:  table doesn't have 'DISTRIBUTED BY' clause, defaulting to distribution columns from LIKE table
CREATE TABLE
testDB=#

4. 插入語句

插入語句 
執行insert語句注意分佈鍵不要爲空,不然分佈鍵默認變成null',數據都被保存到一個節點上會致使分佈不均

insert into test001 values (100,'tom'),(101,'lily'),(102,'jack'),(103,'linda');
insert into test002 values (200,'tom'),(101,'lily'),(202,'jack'),(103,'linda');

5. 更新語句

不能批量對分佈鍵執行update,由於分佈鍵執行update須要將數據重分佈.

testDB=# update test002 set id=203 where id=202;
UPDATE 1
testDB=#

6. 刪除語句delete--truncate

在Greenplum 3.x 的版本中，若是delete 操做涉及子查詢，子查詢的結果還涉及到數據重分佈，這樣的刪除語句會報錯，Greenplum 4.x以上，支持該操做。
testDB=# delete from test001 where name in (select name from test002);
DELETE 4
testDB=# 

若是對整張表執行delete會比較慢，建議使用truncate.

truncate執行truncate直接刪除表的物理文件，而後建立新的數據文件。若是有sql正在操做這張表，那麼truncate會被鎖住，直到表上面的全部鎖會被釋放。

7. 查詢語句

testDB=# select * from test001 x,test002 y where x.id=y.id;
 id  | name  | id  | name  
-----+-------+-----+-------
 103 | linda | 103 | linda
 101 | lily  | 101 | lily
(2 rows)

8. 執行計劃

testDB=# explain select * from test001 x,test002 y where x.id=y.id;
                                  QUERY PLAN                                   
-------------------------------------------------------------------------------
 Gather Motion 6:1  (slice1; segments: 6)  (cost=0.00..862.00 rows=5 width=18)
   ->  Hash Join  (cost=0.00..862.00 rows=1 width=18)
         Hash Cond: (test001.id = test002.id)
         ->  Seq Scan on test001  (cost=0.00..431.00 rows=1 width=9)
         ->  Hash  (cost=431.00..431.00 rows=1 width=9)
               ->  Seq Scan on test002  (cost=0.00..431.00 rows=1 width=9)
 Optimizer: Pivotal Optimizer (GPORCA)
(7 rows)